CN117982107A - Sleep quality monitoring method based on radar signals - Google Patents
Sleep quality monitoring method based on radar signals Download PDFInfo
- Publication number
- CN117982107A CN117982107A CN202410396410.6A CN202410396410A CN117982107A CN 117982107 A CN117982107 A CN 117982107A CN 202410396410 A CN202410396410 A CN 202410396410A CN 117982107 A CN117982107 A CN 117982107A
- Authority
- CN
- China
- Prior art keywords
- data
- degree
- respiratory rate
- sleep
- day
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 75
- 230000003860 sleep quality Effects 0.000 title claims abstract description 42
- 238000012544 monitoring process Methods 0.000 title claims abstract description 36
- 230000036387 respiratory rate Effects 0.000 claims abstract description 125
- 230000007958 sleep Effects 0.000 claims abstract description 77
- 230000002159 abnormal effect Effects 0.000 claims abstract description 59
- 230000036391 respiratory frequency Effects 0.000 claims abstract description 34
- 230000035565 breathing frequency Effects 0.000 claims description 49
- 238000004891 communication Methods 0.000 claims description 43
- 230000008569 process Effects 0.000 claims description 37
- 230000036385 rapid eye movement (rem) sleep Effects 0.000 claims description 22
- 230000000739 chaotic effect Effects 0.000 claims description 18
- 230000003044 adaptive effect Effects 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000012937 correction Methods 0.000 claims description 3
- 230000036578 sleeping time Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 4
- 230000008859 change Effects 0.000 abstract description 3
- 238000012545 processing Methods 0.000 abstract description 2
- 238000005516 engineering process Methods 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 230000036541 health Effects 0.000 description 3
- 230000008667 sleep stage Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000029058 respiratory gaseous exchange Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention relates to the technical field of electric digital data processing, in particular to a sleep quality monitoring method based on radar signals, which comprises the following steps: obtaining sleep data, respiratory rate data and respiratory rate of a user, obtaining the effectiveness of the abnormal respiratory rate data in the respiratory rate data of each day according to the respiratory rate value corresponding to each abnormal respiratory rate data and the number of the abnormal respiratory rate data, and obtaining the confusion degree of extreme points on the respiratory rate data curve of each day according to the effectiveness of the abnormal respiratory rate data in the respiratory rate data of each day; obtaining the similarity between the respiratory frequency data of different days according to the degree of confusion of the extreme points on the fitting curve of the respiratory frequency data of each day, and obtaining the splitting threshold value of each splitting according to the similarity between the respiratory frequency data of different days; and carrying out sleep quality monitoring through clustering according to the splitting threshold value of each splitting. The invention improves the clustering analysis effect by analyzing the sleep data change.
Description
Technical Field
The invention relates to the technical field of electric digital data processing, in particular to a sleep quality monitoring method based on radar signals.
Background
In recent years, attention of people to sleep quality and health is improved, a sleep quality monitoring technology based on radar signals is widely applied and researched in the fields of intelligent wearable equipment and health monitoring, more and more intelligent bracelets, intelligent watches and sleep monitoring equipment begin to integrate sleep quality monitoring functions, and development of the technology is hopeful to provide more accurate and convenient sleep monitoring experience for users, so that the sleep quality is improved and health is promoted.
In the sleep quality monitoring process based on radar signals, similar sleep data are clustered into one type by clustering the sleep data, and then the sleep quality is monitored according to a clustering result. CABDDCG (Clustering Algorithm Based on Dynamic Division of Connected Graph, chinese translation is a clustering algorithm based on dynamic splitting of the connected graph) is a commonly used clustering algorithm, which constructs the connected graph by calculating the similarity between data, then splits the connected graph, and the final split result is the final clustering result. In the traditional method, a fixed splitting threshold is set for splitting in the splitting process of the communication graph, but in the continuous splitting process of the communication graph, the number of nodes and the number of edges in the communication graph are continuously changed, so that the relation among the nodes is also continuously changed, and the fixed splitting threshold is used for controlling the splitting of the communication graph to cause the generation of mistakes in variable data, so that the clustering effect is poor.
Disclosure of Invention
The invention provides a sleep quality monitoring method based on radar signals, which aims to solve the existing problems.
The sleep quality monitoring method based on radar signals adopts the following technical scheme:
one embodiment of the invention provides a radar signal-based sleep quality monitoring method, which comprises the following steps:
acquiring sleep data, respiratory rate data and respiratory rate of a user;
obtaining abnormal respiratory rate data according to the respiratory rate data; obtaining the effectiveness of the abnormal breathing frequency data in the breathing frequency data of each day according to the breathing frequency value corresponding to each abnormal breathing frequency data and the number of the abnormal breathing frequency data;
Obtaining a fluctuation curve of the respiratory rate data of each day according to the effectiveness degree of the abnormal respiratory rate data in the respiratory rate data of each day; obtaining the degree of confusion of the extreme points on the daily respiratory frequency data curve according to the respiratory frequency value of each extreme point on the frequency data curve;
fitting the degree of confusion of the extreme points on the curve according to the breathing frequency data of each day to obtain a confusion degree array; obtaining the similarity between the respiratory rate data of different days according to the effectiveness degree of the abnormal respiratory rate data in the respiratory rate data of each day and each chaotic degree array of each day;
Taking sleep data of a user as nodes and constructing a communication graph; according to the similarity among the respiratory frequency data of different days, the association degree between the nodes after each split is obtained; according to the association degree between the nodes after each split, obtaining a difference value of the association degree between the nodes in the communication in the adjacent split process in the communication split process; obtaining a splitting threshold value of each splitting according to a difference value of association degrees between nodes in the communication in the adjacent splitting process in the communication splitting process;
and carrying out sleep quality monitoring through clustering according to the splitting threshold value of each splitting.
Further, the step of obtaining abnormal respiratory rate data according to respiratory rate data comprises the following specific steps:
abnormal respiratory rate data in the daily respiratory rate data of the user is acquired by using an isolated forest algorithm.
Further, according to the respiratory rate value and the number of the abnormal respiratory rate data corresponding to each abnormal respiratory rate data, the effectiveness of the abnormal respiratory rate data in the respiratory rate data of each day is obtained, and the method comprises the following specific steps:
And recording the sum of the respiratory rate values corresponding to all abnormal respiratory rate data in any day and the absolute value of the difference between the maximum value of the preset theoretical respiratory rate as the effectiveness of the abnormal respiratory rate data in the day.
Further, according to the effectiveness of the abnormal respiratory rate data in the respiratory rate data of each day, a fluctuation curve of the respiratory rate data of each day is obtained, which comprises the following specific steps:
And using the effectiveness of abnormal respiratory rate data in the respiratory rate data of each day as a weight coefficient, and using a polynomial fitting method to fit the respiratory rate data of each day to obtain a fitting curve.
Further, according to the respiratory rate value of each extreme point on the frequency data curve, the degree of confusion of the extreme point on the daily respiratory rate data curve is obtained, and the specific formula is as follows:
Wherein, Represents the/>Person, 5/>Person and/>Degree of confusion of extreme points/>Represents the/>Respiratory rate values corresponding to extreme points,/>Represents the/>Respiratory rate values corresponding to extreme points,/>Represent the firstRespiratory rate values corresponding to extreme points,/>Represents the/>Respiratory rate values corresponding to extreme points,/>Representing absolute value functions,/>Is an exponential function with a base of natural constant.
Further, the specific calculation formula of the similarity between the different days of respiratory frequency data is as follows:
Wherein, Represents the/>Respiratory rate data and the/>Similarity of the respiratory frequency data of days,/>Represents the/>Validity of abnormal respiratory rate data in the respiratory rate data of day,/>Represents the/>Validity of abnormal respiratory rate data in the respiratory rate data of day,/>Represents the/>The first/>, in the array of degree of confusion of the dayA value of degree of confusion,/>Represent the firstThe first/>, in the array of degree of confusion of the dayA value of degree of confusion,/>Represents the/>Number of confusion degree values in the confusion degree array of the day,/>Represents the/>Number of confusion degree values in the confusion degree array of the day,/>Represents the/>Number of REM sleep stages experienced by the patient during the course of day of sleep,/>Represents the/>Number of REM sleep stages the patient experiences during the course of the day's sleep.
Further, the number of REM sleep stages experienced by the patient during the sleep process is specifically obtained by:
each confusion degree value in the confusion degree array is recorded as The average value of all the confusion degree values in the confusion degree array is recorded as/>Traversing from the first confusion degree value in the confusion degree array to find the first value larger than the first valueAnd marking, and continuing to traverse the next confusion degree value to calculate whether the confusion degree value is larger than/>If the target point is larger than the target point, marking the target point as a target point, if the target point is smaller than the target point, marking the target point as/>, and marking the proportion of the target points asWherein/>Representing the number of total clutter level values traversed,/>Representing the number of marked target points, presetting a cut-off threshold value, if/>If the value of (1) is smaller than the cut-off threshold value for the first time, traversing is stopped to obtain a section in the chaotic degree array, and traversing is continued to find the first section larger than/>, taking the stopped chaotic degree value as a starting pointAnd (3) marking the chaotic degree value of the window, continuing to traverse until the last chaotic degree value in the chaotic degree array is traversed, obtaining a plurality of intervals after the traversing is finished, and taking the number of the intervals as the number of REM sleep stages experienced in the sleeping time of the day.
Further, the sleep data of the user is taken as a node and a communication graph is constructed; according to the similarity among the respiratory frequency data of different days, the association degree between the nodes after each split is obtained; according to the association degree between the nodes after each split, obtaining a difference value of the association degree between the nodes in the communication in the adjacent split process in the communication split process, comprising the following specific steps:
taking daily sleep data of a user as a node respectively, constructing a connected graph of all sleep data by using CABDDCG clustering method, and measuring distance by adopting the sleep data of different days of the user Distance, constructing to obtain a plurality of sub-communication graphs;
Wherein, Represents the/>Degree of association between nodes in post-sub-split connected graph and/>,/>Represents the/>Respiratory rate data and the/>Similarity of the respiratory frequency data of days,/>Representing the degree of association between nodes in an initial connected graph,/>Days representing sleep data counted by the user;
Will be the first Degree of association between nodes and/>, in post-sub-split connected graphThe difference of the degree of association between nodes in the post-split connected graph is denoted as/>Post-sub-division and/>And (3) correlation degree difference values among nodes in the sub-split connected graph.
Further, the method for obtaining the splitting threshold value of each splitting according to the difference value of the association degree between the nodes in the communication in the adjacent splitting process in the communication splitting process comprises the following specific steps:
Wherein, Represents the/>Adaptive splitting threshold during sub-splitting,/>Represents the/>Adaptive splitting threshold in word splitting process,/>Representing an initial split threshold of a connected graph,/>Represents the/>Correction direction value of self-adaptive threshold value in sub-splitting process,/>Represents the/>Post-sub-division and/>Relevance degree difference value between nodes in post-sub-split connected graph,/>Representing an absolute value function.
Further, the sleep quality monitoring is performed by clustering according to the splitting threshold value of each splitting, and the method comprises the following specific steps:
Completing the splitting process of the sub-communication graphs by utilizing CABDDCG clustering algorithm according to the splitting threshold value of each splitting in the splitting process of each sub-communication graph to obtain a plurality of clusters, and monitoring the sleep quality of the user according to the number of sleep data in each cluster and the characteristics of the sleep data; presetting a judgment threshold and a proportion threshold;
When the ratio of the number of sleep data to all data amounts in one cluster is larger than a proportional threshold, and the average value of the times of REM sleep stages of all sleep data in the cluster is smaller than a judgment threshold, the sleep quality is good;
When the ratio of the number of sleep data to all data amounts in each cluster is smaller than or equal to a proportional threshold, or the average value of the number of REM sleep stages of all sleep data in the cluster is larger than or equal to a judgment threshold, the sleep quality is poor.
The technical scheme of the invention has the beneficial effects that: CABDDCG is a common clustering method, firstly, a data set is constructed as a connected graph, then a fixed threshold is set as a splitting condition of the connected graph, and then the connected graph is split, the final splitting result is the final clustering result, but the association degree of nodes in the connected graph at different splitting stages in the splitting process of the connected graph is different, so that the splitting result is inaccurate due to the fact that the unified threshold is set, and for CABDDCG algorithm, one-day sleep data of a user is one node. In the embodiment, firstly, the similarity between different nodes is calculated according to the characteristics of the sleep data, and then the splitting threshold value in the splitting process is self-adapted according to the association degree between the nodes in the communication graph of different splitting stages.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flow chart of steps of a sleep quality monitoring method based on radar signals.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following detailed description is given below of the specific implementation, structure, characteristics and effects of the radar signal-based sleep quality monitoring method according to the invention in combination with the accompanying drawings and the preferred embodiment. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the sleep quality monitoring method based on radar signals provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart illustrating steps of a radar signal-based sleep quality monitoring method according to an embodiment of the present invention is shown, the method includes the steps of:
step S001: sleep data, respiratory rate data, and respiratory rate of a user are acquired.
The purpose of this embodiment is to obtain sleep data of a user through radar signals and perform quality detection, and after fourier transformation of the radar data, frequency data corresponding to breathing frequency is used as breathing frequency data of the user, and the breathing frequency data is used to represent the sleep data of the user.
The radar data of the past 2 months of the user is acquired by using the sleep detection radar, in the embodiment, the radar data is acquired once every minute, the radar data is subjected to Fourier transform to acquire the breathing frequency data of the user, the breathing frequency data corresponds to one breathing frequency, namely the breathing frequency of the user for one minute, the theoretical breathing frequency of a normal adult is usually 12 to 20 times per minute, and the maximum value of the theoretical breathing frequency is preset in the embodimentThe method for obtaining the respiratory rate data by performing fourier transform on the radar data is not described in detail in this embodiment, and the sleep data is a time sequence curve when all respiratory rate data in one day is formed into a piece of sleep data of the user.
To this end, sleep data, respiratory rate data and respiratory rate of the user are obtained.
Step S002: obtaining abnormal respiratory rate data according to the respiratory rate data; and obtaining the effectiveness of the abnormal breathing frequency data in the breathing frequency data of each day according to the breathing frequency value corresponding to each abnormal breathing frequency data and the number of the abnormal breathing frequency data.
It should be noted that, taking the sleep data of a user in one day as a node, it includes the respiratory frequency data and the heart rate data of the same day, usually the sleep of a person will go through a plurality of sleep cycles, one sleep cycle generally includes four stages of shallow sleep, medium sleep, deep sleep and REM sleep, because the respiratory frequency values of the first three stages of the four stages of shallow sleep, medium sleep and deep sleep are slowly reduced, and the fluctuation of the respiratory frequency is relatively small, and the respiratory frequency fluctuation in REM sleep state is close to the awake state, so the respiratory frequency fluctuation will be changed from stable to unstable, the respiratory frequency value will gradually rise from stable state to a respiratory frequency interval similar to that of the shallow sleep stage, and unstable fluctuation will be carried out, so the fluctuation of the respiratory frequency in REM sleep state is relatively large, and the user will wake halfway due to the condition of night, dream during sleep, and the like, if the user needs to reentry the sleep cycle, the quality of sleep is affected, so the number of different days of data of the user is obtained through the number of awakenings and the sleep cycle experienced during the sleep process.
Specifically, in the user's firstFor example, the respiratory rate data of the day is used to obtain the user's/>, using an isolated forest algorithmAbnormal respiratory rate data in the respiratory rate data of the day, the abnormal detection of the isolated forest algorithm is performed in the prior art, and the embodiment is not repeated; since the user does not have to wake halfway, the calculated abnormal breathing data may be breathing frequency data of REM sleep stage, and thus the validity of the abnormal breathing frequency data needs to be calculated, and the first/>The calculation formula of the validity of the abnormal respiratory rate data in the respiratory rate of days is as follows:
Wherein, Represents the/>Validity of abnormal respiratory rate data in the respiratory rate data of day,/>Represents the/>Respiratory rate values corresponding to the abnormal respiratory rate data,/>Representing the number of abnormal respiratory rate data,/>Represents the theoretical maximum respiratory rate for four phases of the sleep cycle,/>The larger the value of (2) is, the more/>The greater the validity of the abnormal breathing frequency data, the greater the ratio ofRepresenting an absolute value function.
According to the above manner, the validity of the abnormal respiratory rate data in the respiratory rate data of each day is obtained.
To this end, the validity of the abnormal respiratory rate data in the respiratory rate data of each day is obtained.
Step S003: obtaining a fluctuation curve of the respiratory rate data of each day according to the effectiveness degree of the abnormal respiratory rate data in the respiratory rate data of each day; and obtaining the degree of confusion of the extreme points on the daily respiratory frequency data curve according to the respiratory frequency value of each extreme point on the frequency data curve.
Specifically, according to the effectiveness degree of the abnormal breathing frequency data in the breathing frequency data of each day, a fluctuation curve of the breathing frequency data of each day is obtained, the effectiveness degree of the abnormal breathing frequency data in the breathing frequency data of each day is used as a weight coefficient, and a polynomial fitting method is used for fitting the breathing frequency data of each day to obtain a fitting curve, wherein the polynomial fitting method is a known technology, and the embodiment is not repeated; detecting extreme points of the fitting curve to obtain a plurality of extreme points on the fitting curve, namely the firstThe calculation formula of the degree of confusion of the extreme points on the fitting curve of the breathing frequency data of the day is as follows:
Wherein, Represents the/>Person, 5/>Person and/>Degree of confusion of extreme points/>Represents the/>Respiratory rate values corresponding to extreme points,/>Represents the/>Respiratory rate values corresponding to extreme points,/>Represent the firstRespiratory rate values corresponding to extreme points,/>Represents the/>The respiratory rate values corresponding to the extreme points,The fluctuation of the three extreme points is represented, and the larger the value thereof is, the larger the fluctuation of the three extreme points is represented,/>Representing absolute value functions, while/>The larger the value of (c) is, the greater the degree of deviation of the starting extremum point and the ending extremum point is, the greater the degree of confusion of the corresponding three extremum points is, and the greater the degree of confusion of the corresponding three extremum points isThe present embodiment uses/>, as an exponential function based on natural constantsThe nonlinear growth relation is presented, and the stronger the fluctuation of the extreme points is, the faster the chaotic degree of the extreme points is increased.
According to the mode, the degree of confusion of every adjacent three extreme points on the breathing frequency data fitting curve of each day is obtained.
Thus, the degree of confusion of every adjacent three extreme points on the breathing frequency data fitting curve of each day is obtained.
Step S004: fitting the degree of confusion of the extreme points on the curve according to the breathing frequency data of each day to obtain a confusion degree array; and obtaining the similarity between the respiratory rate data of different days according to the effectiveness degree of the abnormal respiratory rate data in the respiratory rate data of each day and each chaotic degree array of each day.
It should be noted that the degree of confusion of the extreme points in the first three stages of the sleep cycle is relatively small, so the first can be determined by analyzing the change of the degree of confusionREM sleep stage intervals in the day's respiratory rate data.
Specifically, the degree of confusion of all extreme points on the respiratory rate data fitting curve of each day is formed into an array of degree of confusion values, the degree of confusion is recorded as an array of degree of confusion, and each degree of confusion in the array of degree of confusion is recorded asThe average value of all the confusion degree values in the confusion degree array is recorded as/>Traversing from the first confusion degree value in the confusion degree array to find the first value larger than/>And marking, and continuing to traverse the next confusion degree value to calculate whether the confusion degree value is larger than/>If the target point is larger than the target point, marking the target point as a target point, if the target point is smaller than the target point, marking the target point as/>, and marking the proportion of the target points asWherein/>Representing the number of total clutter level values traversed,/>The number of marked target points is represented, a cut-off threshold is preset, and the cut-off threshold in the embodiment adopts/>To describe, if/>If the value of (1) is smaller than the cut-off threshold value for the first time, traversing is stopped to obtain a section in the chaotic degree array, and traversing is continued to find the first section larger than/>, taking the stopped chaotic degree value as a starting pointAccording to the method, stopping until the last chaotic degree value in the chaotic degree array is traversed, obtaining a plurality of intervals after the traversing is finished, taking the number of the intervals as the number of REM sleep stages experienced in the sleeping time of the day, wherein the number of REM sleep stages can be used for representing the sleeping quality of a user, and then/>Respiratory rate data and the/>The formula for calculating the similarity of the respiratory rate data of the day is as follows:
Wherein, Represents the/>Respiratory rate data and the/>Similarity of the respiratory frequency data of days,/>Represents the/>Validity of abnormal respiratory rate data in the respiratory rate data of day,/>Represents the/>Validity of abnormal respiratory rate data in the respiratory rate data of day,/>Represents the/>The first/>, in the array of degree of confusion of the dayA value of degree of confusion,/>Represent the firstThe first/>, in the array of degree of confusion of the dayA value of degree of confusion,/>Represents the/>Number of confusion degree values in the confusion degree array of the day,/>Represents the/>Number of confusion degree values in the confusion degree array of the day,/>Represents the/>Degree of confusion in the respiratory rate data of day,/>Represents the/>Degree of confusion in the respiratory rate data of day,/>Represents the/>Number of REM sleep stages experienced by the patient during the course of day of sleep,/>Represents the/>Number of REM sleep stages the patient experiences during the course of the day's sleep.
The smaller the value of (2) is, the more/>Tianhe/>The higher the similarity of abnormal respiratory rate of the day's respiratory rate data, the simultaneously/>The smaller the value of/>The smaller the value of (2) is, the more/>Tianhe/>The stronger the overall similarity of the respiratory rate data for the day.
According to the above mode, the similarity between the respiratory frequency data of different days is obtained.
Thus, the similarity between the respiratory frequency data of different days is obtained.
Step S005: taking sleep data of a user as nodes and constructing a communication graph; according to the similarity among the respiratory frequency data of different days, the association degree between the nodes after each split is obtained; according to the association degree between the nodes after each split, obtaining a difference value of the association degree between the nodes in the communication in the adjacent split process in the communication split process; and obtaining a splitting threshold value of each splitting according to the difference value of the association degree between nodes in the communication in the adjacent splitting process in the communication splitting process.
It should be noted that, the similarity between nodes is obtained, then a connected graph is constructed according to the similarity value between the nodes, and in the dynamic splitting process of the connected graph, the correlation degree between the nodes in the connected graph to be split and the last connected graph to be split is based on the connected graph to be split.
Specifically, each day of sleep data of a user is used as a node, all the communication graphs of the sleep data are constructed by using the existing method for constructing the communication graphs in CABDDCG clustering method, and distance measurement is adopted between the sleep data of different days of the userThe distance, CABDDCG clustering method is a known technology, and this embodiment is not repeated, and a plurality of sub-connected graphs are constructed and obtained, and a final clustering result can be obtained by performing splitting operation on each sub-connected graph, so that a calculation formula of the association degree between the nodes is as follows:
Wherein, Represents the/>Correlation degree between nodes in post-split connected graph,/>,/>Represent the firstRespiratory rate data and the/>Similarity of the respiratory frequency data of days,/>Representing the degree of association between nodes in an initial connected graph,/>Days of sleep data representing user statistics.
Wherein,Represents the/>Post-sub-division and/>Relevance degree difference value between nodes in post-sub-split connected graph,/>Represents the/>Correlation degree between nodes in post-split connected graph,/>Represents the/>The association degree between nodes in the post-split connected graph.
Further, a split threshold is obtained according to the association degree between the nodes, and a calculation formula of the split threshold is as follows:
Wherein, Represents the/>Adaptive splitting threshold during sub-splitting,/>Represents the/>Adaptive splitting threshold in word splitting process,/>Representing an initial split threshold of a connected graph,/>Represents the/>Correction direction value of self-adaptive threshold value in sub-splitting process,/>Represents the/>Post-sub-division and/>Relevance degree difference value between nodes in post-sub-split connected graph,/>Representing an absolute value function.
Wherein,When/>When it indicates the/>The association degree between nodes in the sub-split connected graph is larger than that of the/>Degree of association from node to node in a sub-split connected graph, therefore the/>The splitting threshold of (2) is at the/>The sub-splitting threshold is increased by a bit on the basis of the sub-splitting threshold, because the higher the similarity between nodes in the communication graph is, the higher the density in the communication graph is, the higher the splitting threshold is required to ensure the stability of the communication graph, and splitting of the communication graph is avoided because of the change of a bit,/>When (1)The sub-splitting threshold is atThe sub-split threshold continues to be reduced based on it.
To this end, a split threshold for each split is obtained.
Step S006: and carrying out sleep quality monitoring through clustering according to the splitting threshold value of each splitting.
Completing the splitting process of the sub-communication graphs by utilizing CABDDCG clustering algorithm according to the splitting threshold value of each splitting in the splitting process of each sub-communication graph to obtain a plurality of clusters, and monitoring the sleep quality of the user according to the number of sleep data in each cluster and the characteristics of the sleep data;
the judgment threshold and the proportion threshold are preset, the judgment threshold of the embodiment is described by 5, and the proportion threshold is described by Other values may be set in other embodiments as described.
When the ratio of the number of sleep data to all data amounts in one cluster is larger than a proportional threshold value and the average value of the times of REM sleep stages of all sleep data in the cluster is smaller than a judgment threshold value, the sleep quality is good;
And when the ratio of the number of sleep data to all data amounts in each cluster is smaller than or equal to a proportional threshold, or the average value of the number of REM sleep stages of all sleep data in the cluster is larger than or equal to a judgment threshold, the sleep quality is poor.
This embodiment is completed.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the invention, but any modifications, equivalent substitutions, improvements, etc. within the principles of the present invention should be included in the scope of the present invention.
Claims (10)
1. The sleep quality monitoring method based on the radar signal is characterized by comprising the following steps of:
acquiring sleep data, respiratory rate data and respiratory rate of a user;
obtaining abnormal respiratory rate data according to the respiratory rate data; obtaining the effectiveness of the abnormal breathing frequency data in the breathing frequency data of each day according to the breathing frequency value corresponding to each abnormal breathing frequency data and the number of the abnormal breathing frequency data;
Obtaining a fluctuation curve of the respiratory rate data of each day according to the effectiveness degree of the abnormal respiratory rate data in the respiratory rate data of each day; obtaining the degree of confusion of the extreme points on the daily respiratory frequency data curve according to the respiratory frequency value of each extreme point on the frequency data curve;
fitting the degree of confusion of the extreme points on the curve according to the breathing frequency data of each day to obtain a confusion degree array; obtaining the similarity between the respiratory rate data of different days according to the effectiveness degree of the abnormal respiratory rate data in the respiratory rate data of each day and each chaotic degree array of each day;
Taking sleep data of a user as nodes and constructing a communication graph; according to the similarity among the respiratory frequency data of different days, the association degree between the nodes after each split is obtained; according to the association degree between the nodes after each split, obtaining a difference value of the association degree between the nodes in the communication in the adjacent split process in the communication split process; obtaining a splitting threshold value of each splitting according to a difference value of association degrees between nodes in the communication in the adjacent splitting process in the communication splitting process;
and carrying out sleep quality monitoring through clustering according to the splitting threshold value of each splitting.
2. The radar signal-based sleep quality monitoring method as claimed in claim 1, wherein the obtaining abnormal breathing frequency data according to the breathing frequency data comprises the following specific steps:
abnormal respiratory rate data in the daily respiratory rate data of the user is acquired by using an isolated forest algorithm.
3. The method for monitoring sleep quality based on radar signals according to claim 1, wherein the step of obtaining the validity of the abnormal breathing frequency data in the breathing frequency data of each day according to the breathing frequency value corresponding to each abnormal breathing frequency data and the number of the abnormal breathing frequency data comprises the following specific steps:
And recording the sum of the respiratory rate values corresponding to all abnormal respiratory rate data in any day and the absolute value of the difference between the maximum value of the preset theoretical respiratory rate as the effectiveness of the abnormal respiratory rate data in the day.
4. The radar signal-based sleep quality monitoring method according to claim 1, wherein the step of obtaining the fluctuation curve of the breathing frequency data of each day according to the effectiveness of the abnormal breathing frequency data in the breathing frequency data of each day comprises the following specific steps:
And using the effectiveness of abnormal respiratory rate data in the respiratory rate data of each day as a weight coefficient, and using a polynomial fitting method to fit the respiratory rate data of each day to obtain a fitting curve.
5. The radar signal-based sleep quality monitoring method according to claim 1, wherein the obtaining the degree of confusion of the extreme points on the daily respiratory frequency data curve according to the respiratory frequency value of each extreme point on the frequency data curve comprises the following specific formulas:
Wherein, Represents the/>Person, 5/>Person and/>Degree of confusion of extreme points/>Represents the/>Respiratory rate values corresponding to extreme points,/>Represents the/>Respiratory rate values corresponding to extreme points,/>Represents the/>Respiratory rate values corresponding to extreme points,/>Represents the/>Respiratory rate values corresponding to extreme points,/>Representing absolute value functions,/>Is an exponential function with a base of natural constant.
6. The radar signal-based sleep quality monitoring method according to claim 1, wherein the specific calculation formula of the similarity between the different days of respiratory frequency data is as follows:
Wherein, Represents the/>Respiratory rate data and the/>Similarity of the respiratory frequency data of days,/>Represents the/>Validity of abnormal respiratory rate data in the respiratory rate data of day,/>Represents the/>Validity of abnormal respiratory rate data in the respiratory rate data of day,/>Represents the/>The first/>, in the array of degree of confusion of the dayA value of degree of confusion,/>Represents the/>The first/>, in the array of degree of confusion of the dayA value of degree of confusion,/>Represents the/>Number of confusion degree values in the confusion degree array of the day,/>Represents the/>Number of confusion degree values in the confusion degree array of the day,/>Represents the/>Number of REM sleep stages experienced by the patient during the course of day of sleep,/>Represents the/>Number of REM sleep stages the patient experiences during the course of the day's sleep.
7. The radar signal-based sleep quality monitoring method according to claim 6, wherein the number of REM sleep stages experienced by the patient during sleep is specifically obtained by:
each confusion degree value in the confusion degree array is recorded as The average value of all the confusion degree values in the confusion degree array is recorded as/>Traversing from the first confusion degree value in the confusion degree array to find the first value larger than/>And marking, and continuing to traverse the next confusion degree value to calculate whether the confusion degree value is larger than/>If the target point is larger than the target point, marking the target point as a target point, if the target point is smaller than the target point, marking the target point as/>, and marking the proportion of the target points asWherein/>Representing the number of total clutter level values traversed,/>Representing the number of marked target points, presetting a cut-off threshold value, if/>If the value of (1) is smaller than the cut-off threshold value for the first time, traversing is stopped to obtain a section in the chaotic degree array, and traversing is continued to find the first section larger than/>, taking the stopped chaotic degree value as a starting pointAnd (3) marking the chaotic degree value of the window, continuing to traverse until the last chaotic degree value in the chaotic degree array is traversed, obtaining a plurality of intervals after the traversing is finished, and taking the number of the intervals as the number of REM sleep stages experienced in the sleeping time of the day.
8. The radar signal-based sleep quality monitoring method according to claim 1, wherein the sleep data of the user is used as a node and a connectivity graph is constructed; according to the similarity among the respiratory frequency data of different days, the association degree between the nodes after each split is obtained; according to the association degree between the nodes after each split, obtaining a difference value of the association degree between the nodes in the communication in the adjacent split process in the communication split process, comprising the following specific steps:
taking daily sleep data of a user as a node respectively, constructing a connected graph of all sleep data by using CABDDCG clustering method, and measuring distance by adopting the sleep data of different days of the user Distance, constructing to obtain a plurality of sub-communication graphs;
Wherein, Represents the/>Degree of association between nodes in post-sub-split connected graph and/>,/>Represents the/>Respiratory rate data and the/>Similarity of the respiratory frequency data of days,/>Representing the degree of association between nodes in an initial connected graph,/>Days representing sleep data counted by the user;
Will be the first Degree of association between nodes and/>, in post-sub-split connected graphThe difference of the degree of association between nodes in the post-split connected graph is denoted as/>Post-sub-division and/>And (3) correlation degree difference values among nodes in the sub-split connected graph.
9. The sleep quality monitoring method based on radar signals according to claim 1, wherein the obtaining the splitting threshold value of each splitting according to the difference value of the association degree between the nodes in the communication in the adjacent splitting in the communication splitting comprises the following specific steps:
Wherein, Represents the/>Adaptive splitting threshold during sub-splitting,/>Represents the/>Adaptive splitting threshold in word splitting process,/>Representing an initial split threshold of a connected graph,/>Represents the/>Correction direction value of self-adaptive threshold value in sub-splitting process,/>Represents the/>Post-sub-division and/>Relevance degree difference value between nodes in post-sub-split connected graph,/>Representing an absolute value function.
10. The method for monitoring sleep quality based on radar signals according to claim 1, wherein the step of monitoring sleep quality by clustering according to the splitting threshold value of each splitting comprises the following specific steps:
Completing the splitting process of the sub-communication graphs by utilizing CABDDCG clustering algorithm according to the splitting threshold value of each splitting in the splitting process of each sub-communication graph to obtain a plurality of clusters, and monitoring the sleep quality of the user according to the number of sleep data in each cluster and the characteristics of the sleep data; presetting a judgment threshold and a proportion threshold;
When the ratio of the number of sleep data to all data amounts in one cluster is larger than a proportional threshold, and the average value of the times of REM sleep stages of all sleep data in the cluster is smaller than a judgment threshold, the sleep quality is good;
When the ratio of the number of sleep data to all data amounts in each cluster is smaller than or equal to a proportional threshold, or the average value of the number of REM sleep stages of all sleep data in the cluster is larger than or equal to a judgment threshold, the sleep quality is poor.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410396410.6A CN117982107B (en) | 2024-04-03 | 2024-04-03 | Sleep quality monitoring method based on radar signals |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410396410.6A CN117982107B (en) | 2024-04-03 | 2024-04-03 | Sleep quality monitoring method based on radar signals |
Publications (2)
Publication Number | Publication Date |
---|---|
CN117982107A true CN117982107A (en) | 2024-05-07 |
CN117982107B CN117982107B (en) | 2024-05-28 |
Family
ID=90893585
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410396410.6A Active CN117982107B (en) | 2024-04-03 | 2024-04-03 | Sleep quality monitoring method based on radar signals |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117982107B (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020109863A2 (en) * | 2018-11-30 | 2020-06-04 | Mari Co., Ltd | Sleep apnea treatment apparatus and method |
CN112716474A (en) * | 2021-01-20 | 2021-04-30 | 复旦大学 | Non-contact sleep state monitoring method and system based on biological microwave radar |
CN115778352A (en) * | 2023-02-02 | 2023-03-14 | 北京华声医学科技开发公司 | Sleep quality evaluation method, device, system and medium based on millimeter wave radar |
KR20230133202A (en) * | 2022-03-10 | 2023-09-19 | 주식회사 비트센싱 | Server, method and computer program for analyzing of sleep time using radar |
CN117064410A (en) * | 2023-08-28 | 2023-11-17 | 北京京航计算通讯研究所 | Military physical training risk assessment method and system based on deep learning |
CN117788459A (en) * | 2024-02-23 | 2024-03-29 | 青岛正大正电力环保设备有限公司 | Chain abrasion detection method based on image characteristics |
-
2024
- 2024-04-03 CN CN202410396410.6A patent/CN117982107B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020109863A2 (en) * | 2018-11-30 | 2020-06-04 | Mari Co., Ltd | Sleep apnea treatment apparatus and method |
CN112716474A (en) * | 2021-01-20 | 2021-04-30 | 复旦大学 | Non-contact sleep state monitoring method and system based on biological microwave radar |
KR20230133202A (en) * | 2022-03-10 | 2023-09-19 | 주식회사 비트센싱 | Server, method and computer program for analyzing of sleep time using radar |
CN115778352A (en) * | 2023-02-02 | 2023-03-14 | 北京华声医学科技开发公司 | Sleep quality evaluation method, device, system and medium based on millimeter wave radar |
CN117064410A (en) * | 2023-08-28 | 2023-11-17 | 北京京航计算通讯研究所 | Military physical training risk assessment method and system based on deep learning |
CN117788459A (en) * | 2024-02-23 | 2024-03-29 | 青岛正大正电力环保设备有限公司 | Chain abrasion detection method based on image characteristics |
Non-Patent Citations (1)
Title |
---|
JUN YOUNG AN: ""Non‑contact diagnosis of sleep breathing disorders using infrared optical gas imaging: a prospective observational study"", 《SCIENTIFIC REPORTS》, 6 December 2022 (2022-12-06) * |
Also Published As
Publication number | Publication date |
---|---|
CN117982107B (en) | 2024-05-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107688553B (en) | Method for detecting electrocardiographic waveform characteristics based on wavelet transform and logistic regression algorithm | |
CN117540238B (en) | Data security management method for industrial digital information acquisition device | |
CN104921736A (en) | Continuous blood glucose monitoring device comprising parameter estimation function filtering module | |
CN117982107B (en) | Sleep quality monitoring method based on radar signals | |
CN113679419B (en) | Adjustable Doppler frequency spectrum envelope parameter calculation method | |
CN114582503A (en) | Blood pressure monitoring method based on multiple linear regression and correlation weighting selection algorithm | |
CN116414076A (en) | Intelligent monitoring system for recovered alcohol production data | |
CN115868941A (en) | Information management method for intelligent ring | |
CN117807551B (en) | Heart rate abnormality capturing method and system based on intelligent ring | |
CN117972357B (en) | Intelligent water level monitoring data processing method of water level measuring device | |
CN117373600B (en) | Medical detection vehicle data optimal storage method | |
CN107835025B (en) | Adaptive compression method and system for historical data of power load | |
CN117093857A (en) | Rapid processing method for intelligent ring test data | |
Benchekroun et al. | Analysis of the impact of inter-beat-interval interpolation on real-time HRV feature estimation for e-health applications | |
US20200330040A1 (en) | Method and system for detecting noise in vital sign signal | |
CN114073519B (en) | Noninvasive blood glucose detection method and device based on heart rate variability | |
WO2019036926A1 (en) | Acceleration information-based foot step counting method and apparatus, and device | |
CN112386250A (en) | Intelligent sleep monitoring operation method | |
TWI829575B (en) | Physiological feature detection method, electronic device and server using thereof | |
CN110025323B (en) | Infant emotion recognition method | |
CN118285762B (en) | Radar signal analysis processing method for sleep monitoring | |
CN117951630B (en) | Abnormal breathing pattern recognition method for patient with severe respiratory disease | |
CN117033906A (en) | BCG signal heart rate extraction method based on multidimensional features | |
CN118039052B (en) | Intelligent management method and system for medical health of user | |
CN116362416A (en) | Regional longan annual output prediction method and device and computer equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |